Scholarlink

<p>The El-Bahariya depression in the Western Desert of Egypt is well-known for its iron ore deposits, with mineralization recorded in five well-known locations. This study is the first integration of hyperspectral PRISMA and aeromagnetic analysis for the Bahariya iron ore deposits. Automated lithological mapping of the study area was performed using machine learning algorithms (MLA) such as Random Forest (RF) and Support Vector Machine (SVM), applied to stacked (ASTER + Sentinel-2) data, achieving an overall accuracy of up to 91.73%, Kappa accuracy of 89.99%, and F1-score of 93.98%. Hyperspectral PRISMA data analysis identified diagnostic absorption features (1.95–2.3&#xa0;μm) associated with hematite, goethite, and limonite. Advanced spectral methods, including Pixel Purity Index (PPI) and Spectral Angle Mapper (SAM), successfully discriminated between hematite and limonite concentrations, as validated by field surveys, an ASD spectroradiometer, and USGS laboratory spectra. Results revealed ferruginous sandstone distributions, iron-rich zones, and previously undetected high-potential mineralization targets. Complementary high-resolution aeromagnetic data, processed using edge detection filters and CET techniques (CET-GA, CET-PA), resolved structural controls on mineralization. Major NE-SW and NW-SE trending lineaments, alongside minor N-S and E-W structures, were identified as fluid conduits for hydrothermal iron oxide emplacement. Euler deconvolution constrained magnetic sources to shallow depths (&lt; 2&#xa0;km), aligning with surface-derived anomalies. High magnetic susceptibility zones correlated strongly with remote sensing-identified iron-rich areas, including known mines, and highlighted unexplored anomalies in the northwestern and central regions. Geochemically, we identified economic ore-grade ironstones distinct (FeO<sup>t</sup> = 26–46 wt%, MnO &lt; 0.11 wt%, P₂O₅ 0.45–0.84 wt%) from non-economic Mn-rich carbonate lenses (MnO 4.85–7.76 wt%) and barren siliceous rocks. The integration of spectral and magnetic datasets confirmed the spatial coherence of mineralization signals, along with the detection of new high-potential zones for iron mineralization, demonstrating the synergistic utility of these datasets in defining exploration targets. The proposed methods were highly effective in mapping iron oxide deposits within the four major zones: Nasser, Gabal-Ghurabi, El-Gadidah, and El-Harrah.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Integrated remote sensing and aeromagnetic datasets for mapping iron mineralization potential in the El-Bahariya depression in the Western Desert of Egypt

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月25日

查看全文
  • Saif M. Abo Khashaba,
  • Safaa M. Hassan,
  • Mohamed F. Sadek,
  • Noureldin Laban,
  • Safwat S. Gabr,
  • Mohamed R. Metwalli,
  • Mabrouk Sami,
  • Suhail S. Alhejji,
  • Ioan V. Sanislav,
  • Mostafa Nagy

摘要

The El-Bahariya depression in the Western Desert of Egypt is well-known for its iron ore deposits, with mineralization recorded in five well-known locations. This study is the first integration of hyperspectral PRISMA and aeromagnetic analysis for the Bahariya iron ore deposits. Automated lithological mapping of the study area was performed using machine learning algorithms (MLA) such as Random Forest (RF) and Support Vector Machine (SVM), applied to stacked (ASTER + Sentinel-2) data, achieving an overall accuracy of up to 91.73%, Kappa accuracy of 89.99%, and F1-score of 93.98%. Hyperspectral PRISMA data analysis identified diagnostic absorption features (1.95–2.3 μm) associated with hematite, goethite, and limonite. Advanced spectral methods, including Pixel Purity Index (PPI) and Spectral Angle Mapper (SAM), successfully discriminated between hematite and limonite concentrations, as validated by field surveys, an ASD spectroradiometer, and USGS laboratory spectra. Results revealed ferruginous sandstone distributions, iron-rich zones, and previously undetected high-potential mineralization targets. Complementary high-resolution aeromagnetic data, processed using edge detection filters and CET techniques (CET-GA, CET-PA), resolved structural controls on mineralization. Major NE-SW and NW-SE trending lineaments, alongside minor N-S and E-W structures, were identified as fluid conduits for hydrothermal iron oxide emplacement. Euler deconvolution constrained magnetic sources to shallow depths (< 2 km), aligning with surface-derived anomalies. High magnetic susceptibility zones correlated strongly with remote sensing-identified iron-rich areas, including known mines, and highlighted unexplored anomalies in the northwestern and central regions. Geochemically, we identified economic ore-grade ironstones distinct (FeOt = 26–46 wt%, MnO < 0.11 wt%, P₂O₅ 0.45–0.84 wt%) from non-economic Mn-rich carbonate lenses (MnO 4.85–7.76 wt%) and barren siliceous rocks. The integration of spectral and magnetic datasets confirmed the spatial coherence of mineralization signals, along with the detection of new high-potential zones for iron mineralization, demonstrating the synergistic utility of these datasets in defining exploration targets. The proposed methods were highly effective in mapping iron oxide deposits within the four major zones: Nasser, Gabal-Ghurabi, El-Gadidah, and El-Harrah.